The long term goal of this research program is to uncover and understand mutational pathways, the DNA repair processes or avoidance mechanisms that counteract these pathways and the consequences of defects in these systems. Our basic approach has often involved characterizing "mutator strains" that have higher mutation rates than wild-type, and then determining the affected pathway or repair system. We have developed new approaches for the detection of mutators, and this has allowed us to define several new mutational pathways, including one that results from the overexpression of the EmrR repressor of a multi-drug resistance efflux pump. Overexpressing a common repressor is one way to detect mutational pathways that back each other up and thus require two knockouts to produce a mutator phenotype. We hypothesize that this system acts to pump out mutagenic products of metabolism and represents the cell's first line of defense against mutagenesis. This proposal seeks to test this idea by further charactering this novel system, and also to characterize other new mutational pathways we have discovered. For instance, the use of metagenomic libraries has revealed that multiple copies of extensively repeated sequences results in global genomic instability. This will be further investigated. We will also utilize the recently completed E. coli knockout collection, consisting of close to 4,000 strains, each carrying an in-frame deletion of one of the orfs in the E. coli genome, together with our own tools to screen for new mutational pathways. We will use gene fusions to study the regulation of different repair genes and to characterize the action of new mutators, and will use several pathogen or extremophile genomes (Bordetella pertussis, Campylobacterjejeuni, and Deinococcus radiodurans) as a source for genes that might provoke mutator phenotypes in E. coli, and also that would complement different E. coli repair defects. We will also construct a system for studying mutagenesis in a pathogen such as B. pertussis. [unreadable] [unreadable] [unreadable]